Code page translation problem

Question

Hi guys,
I have the following scenario in XI:
     1. SAP R/3 creates a file in code page 1208 in AS 400
     2. The file is transferred via XI to the client's AS400 system. In XI this is a simple file-to-file scenario that reads from ftp location where SAP R/3 is found and copies the file into XI file system, it then runs an OS command after message processing (receiver file communication channel), that ftp's the file to the client's AS400.
     3. Our client then runs a translation CLP (I am not sure that local IT fully control this CLP) expecting to have as result the file in the correct CP. The correct CP is 813.
The strange thing is that the ABAP program returns some files with Codepage 1208 and some others with 813. When the codePage is SAP R/3 is 813, the XI transfers correctly to the client's machine and all characters are then displayed just fine.
However, when the codepage in SAP R/3 file system is 1208 then we cannot get the characters correctly in target system.
We could not assume how the codepage changes in SAP R/3 system since the ABAP program remains the same.
In XI I have used all the possible encodings: Text, Binary and all codepages.
Any ideas??

Answer 1

When I understand this right, the ABAP downloads the file in CP1208, then a CLP transfers this to CP813 and afterwards the file is processed by file adapter.
I guess that the file adapter is sometimes too fast and processes the file before it is prepared with the CLP.
Maybe it helps when the CLP renames the file after processing so file adapter and to set wildcards in the file adapter channel for the new name.
When you set file to binara in sender and receiver and you have no mapping, the file will be transferred unchanged.
Regards
Stefan